Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for classifying an input containing data by a computer, the method comprising: receiving, by the computer, a list of categories from a classification system stored at the computer, wherein each category in the list is assigned a corresponding distinct correlation score that indicates a relevance of a particular category to a particular classification, wherein the distinct correlation score for each category is based on an accumulation of scores for a number of category values for the category; selecting, by the computer, a sub-list of categories, wherein the sub-list comprises those categories in the list that have corresponding distinct correlation scores above a predetermined value, and wherein the sub-list of categories comprises less than all of the categories in the list; receiving, by the computer, the input containing the data, wherein the data comprises a snapshot, wherein the data is organized by a plurality of input attributes, wherein each input attribute has a corresponding input category and a corresponding input value, such that the plurality of input attributes has a plurality of input categories and a corresponding plurality of input values, wherein at least some of the plurality of input attributes has corresponding input categories that match at least some of a plurality of categories of the list; generating, by the computer, a truncated snapshot, the truncated snapshot comprising only attributes from the plurality of input attributes that have the corresponding input categories that match categories in the sub-list of categories; and classifying the data, by the computer, using the truncated snapshot and the classification system.
This invention relates to data classification by computers. It addresses the problem of efficiently and accurately classifying input data by reducing the complexity of the data and the classification system. The method involves a computer receiving a list of categories from a stored classification system. Each category has a distinct correlation score indicating its relevance for classification. These scores are derived from accumulating scores of category values associated with each category. The computer then selects a sub-list of categories, including only those with correlation scores exceeding a predetermined threshold, ensuring this sub-list is smaller than the full list. Next, the computer receives input data, structured as a snapshot organized by multiple input attributes. Each attribute has an input category and an input value. Crucially, some input categories must match categories from the classification system's list. Based on this, a truncated snapshot is generated, containing only those input attributes whose categories are present in the selected sub-list of categories. Finally, the computer classifies the original data using this truncated snapshot and the classification system.
2. The method of claim 1 further comprising: selecting, by the computer, the predetermined value by subtracting a first correlation score from a second correlation score to form a difference, and selecting the second correlation score as the predetermined value when the difference exceeds a given value.
This invention relates to a method for selecting a predetermined value in a computer system, particularly for use in correlation-based decision-making processes. The method addresses the challenge of dynamically determining an optimal threshold or reference value based on correlation scores to improve accuracy in systems that rely on comparative analysis. The method involves calculating two correlation scores, which represent the degree of relationship between different datasets or variables. The first correlation score is subtracted from the second correlation score to form a difference. If this difference exceeds a predefined threshold, the second correlation score is selected as the predetermined value. This ensures that the selected value is dynamically adjusted based on the relative strength of the correlations, enhancing the system's adaptability and precision in decision-making tasks. The method is particularly useful in applications where correlation-based thresholds are used for filtering, classification, or optimization, such as in machine learning, signal processing, or data analysis. By automatically adjusting the predetermined value based on correlation differences, the system can better handle variations in input data and improve performance in real-time or adaptive environments. The approach reduces the need for manual tuning and increases the robustness of correlation-driven processes.
3. The method of claim 2 , wherein the list is an ordered list, wherein the predetermined value is selected by comparing the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences in correlation scores.
This invention relates to data analysis and categorization, specifically improving the accuracy of categorization by dynamically adjusting thresholds based on correlation scores. The problem addressed is the challenge of determining optimal thresholds for categorizing data into ordered categories when correlation scores between adjacent categories vary significantly. Traditional methods often use fixed thresholds, which can lead to misclassification when data distribution is uneven. The method involves generating an ordered list of categories and calculating correlation scores between adjacent categories. A predetermined value, or threshold, is selected by comparing the difference in correlation scores between adjacent categories to other differences in the list. The threshold is chosen when the difference exceeds a given percentage change relative to the other differences. This adaptive approach ensures that thresholds are dynamically adjusted based on the data's inherent structure, improving categorization accuracy. The method may also include preprocessing data to remove noise or irrelevant features, as well as validating the selected threshold by testing its performance on a subset of the data. The dynamic threshold selection helps mitigate errors caused by fixed thresholds in non-uniform data distributions, making the categorization process more robust and reliable. This technique is particularly useful in applications like machine learning, data clustering, and decision-making systems where precise categorization is critical.
4. The method of claim 1 further comprising: prior to receiving the data, training the classification system by using known data collected from known observations and instructing the classification system as to which of the known observations falls into a given classification of the classification system.
This invention relates to a method for improving the accuracy of a classification system used to process and categorize data. The system addresses the challenge of accurately classifying observations by leveraging known data to train the classification system before it processes new, unclassified data. The method involves collecting known observations, which are data points with predefined classifications, and using these to train the classification system. During training, the system is instructed on which known observations correspond to which classifications, allowing it to learn patterns and relationships within the data. This pre-training step enhances the system's ability to accurately classify new, unknown data when it is later received. The method ensures that the classification system is properly calibrated and optimized before being deployed, reducing errors and improving reliability in real-world applications. The training process may involve various machine learning techniques, such as supervised learning, where the system iteratively adjusts its classification criteria based on feedback from the known data. This approach is particularly useful in fields like image recognition, natural language processing, and predictive analytics, where accurate classification is critical. By pre-training the system with labeled data, the method ensures that the classification system performs with high accuracy when processing new, unclassified observations.
5. The method of claim 1 , wherein the computer comprises a processor and an associative memory, the associative memory comprising a plurality of data and a plurality of associations among the plurality of data, wherein the plurality of data is collected into associated groups, wherein the associative memory is configured to be queried based on at least indirect relationships among the plurality of data, wherein the data is received in the associative memory, and wherein the step of selecting the sub-list further comprises: the associative memory cycling through all of the plurality of categories, determining whether a given category has a corresponding correlation score above the predetermined value, and including the given category in the sub-list if the corresponding correlation score is above the predetermined value.
This invention relates to a computer system with an associative memory for organizing and querying data based on indirect relationships. The system addresses the challenge of efficiently retrieving relevant data from large datasets by leveraging associative memory structures that store both data and their interconnections. The associative memory contains multiple data entries grouped into associated categories, allowing queries to traverse indirect relationships between data points. When new data is received, the system processes it by cycling through all available categories, evaluating each category's correlation score against a predefined threshold. If a category's score exceeds the threshold, it is included in a filtered sub-list of results. This approach enables dynamic and context-aware data retrieval, improving the accuracy and relevance of search outcomes in applications requiring complex relationship mapping, such as knowledge graphs, recommendation systems, or semantic search engines. The associative memory's ability to handle indirect relationships enhances flexibility in querying, making it suitable for domains where direct connections between data are insufficient for meaningful retrieval.
6. The method of claim 1 , wherein the list is ordered from highest corresponding distinct correlation score to lowest corresponding distinct correlation score.
A method for organizing data entries based on correlation scores involves generating a list of data entries, each associated with a distinct correlation score representing a relationship between the entry and a reference data set. The method calculates these correlation scores by analyzing the data entries and the reference data set to determine their degree of similarity or relevance. The calculated scores are then used to rank the data entries, with the highest correlation scores indicating the strongest relationships. The list of entries is ordered from highest to lowest correlation score, allowing for efficient retrieval of the most relevant entries. This ranking can be applied in various domains, such as search engines, recommendation systems, or data analysis tools, where prioritizing entries by relevance is crucial. The method ensures that the most strongly correlated entries are presented first, improving the efficiency and accuracy of data retrieval and processing.
7. The method of claim 1 , wherein selecting the sub-list is performed at a time selected from the group consisting of: before receiving the input and after receiving the input.
A method for optimizing data processing in a computing system involves selecting a sub-list of data items from a larger list to improve efficiency. The selection process can occur either before or after receiving an input that triggers the data processing operation. This approach reduces computational overhead by narrowing down the dataset early or dynamically adjusting the selection based on the input. The method is particularly useful in systems where processing large datasets is resource-intensive, such as in database queries, machine learning model training, or real-time data analysis. By pre-filtering or post-filtering the data, the system avoids unnecessary computations, leading to faster response times and lower energy consumption. The selection criteria for the sub-list may include relevance, priority, or other metadata attributes, ensuring that the most critical or relevant data is processed first. This method enhances performance in applications requiring rapid data retrieval or analysis, such as search engines, recommendation systems, or financial transaction processing. The flexibility to choose between pre-selection or post-selection allows the system to adapt to different operational constraints and user requirements.
8. The method of claim 1 , wherein the computer selects the predetermined value either by (1) subtracting a first correlation score from a second correlation score to form a difference, and selecting the second correlation score as the predetermined value when the difference exceeds a given value, or (2) selecting the predetermined value by comparing the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences.
This invention relates to a method for selecting a predetermined value in a computer-implemented system that processes correlation scores between categories in a ranked list. The problem addressed is the need for an automated and adaptive approach to determine a threshold or cutoff value in a ranked list based on correlation score differences, ensuring accurate categorization or decision-making. The method involves analyzing correlation scores assigned to adjacent categories in a ranked list. The computer selects the predetermined value by either (1) calculating the difference between a first and a second correlation score, then choosing the second correlation score as the predetermined value if this difference exceeds a predefined threshold, or (2) comparing the difference to other differences between adjacent higher-ranked categories. In the second approach, the predetermined value is selected when the difference exceeds a given percentage change relative to the other differences, ensuring consistency with broader trends in the data. This adaptive selection process allows the system to dynamically adjust the predetermined value based on the distribution of correlation scores, improving the reliability of subsequent decisions or classifications derived from the ranked list. The method ensures that the selected value is statistically significant and contextually relevant within the ranked structure.
9. A data processing system comprising: a processor; a bus connected to the processor; a non-transitory computer readable storage medium connected to the bus, the non-transitory computer readable storage medium storing a computer program product which, when executed by the processor, performs a computer implemented method for classifying an input containing data, the computer program product comprising: computer usable program code for receiving a list of categories from a classification system stored at the computer, wherein each category in the list is assigned a corresponding distinct correlation score that indicates a relevance of a particular category to a particular classification, wherein the distinct correlation score for each category is based on an accumulation of scores for a number of category values for the category; computer usable program code for selecting a sub-list of categories, wherein the sub-list comprises those categories in the list that have corresponding distinct correlation scores above a predetermined value, and wherein the sub-list of categories comprises less than all of the categories in the list; computer usable program code for receiving the input containing the data, wherein the data comprises a snapshot, wherein the data is organized by a plurality of input attributes, wherein each input attribute has a corresponding input category and a corresponding input value, such that the plurality of input attributes has a plurality of input categories and a corresponding plurality of input values, wherein at least some of the plurality of input attributes has corresponding input categories that match at least some of a plurality of categories of the list of categories; computer usable program code for generating a truncated snapshot, the truncated snapshot comprising only attributes from the plurality of input attributes that have corresponding input categories that match categories in the sub-list of categories; and computer usable program code for classifying the data using the truncated snapshot and the classification system.
A data processing system classifies input data by filtering relevant categories to improve efficiency and accuracy. The system includes a processor, a bus, and a non-transitory storage medium storing a computer program. The program receives a list of categories from a classification system, where each category has a distinct correlation score indicating its relevance to classification. The score is derived from accumulated scores of multiple category values. The system selects a sub-list of categories with correlation scores above a predetermined threshold, reducing the total number of categories considered. The system then receives input data organized by multiple attributes, each with a category and value. Some input categories match categories in the original list. The system generates a truncated snapshot containing only attributes whose categories match those in the sub-list. Finally, the system classifies the data using the truncated snapshot and the classification system. This approach enhances classification performance by focusing on the most relevant categories, reducing computational overhead and improving accuracy.
10. The data processing system of claim 9 , wherein the computer program product further comprises: computer usable program code for selecting the predetermined value by subtracting a first correlation score from a second correlation score to form a difference, and selecting the second correlation score as the predetermined value when the difference exceeds a given value.
This invention relates to data processing systems designed to improve decision-making by analyzing correlation scores between data elements. The system addresses the challenge of accurately determining a predetermined value used in data processing operations, particularly in scenarios where correlation scores between different data elements need to be compared and evaluated. The system includes a computer program product with executable code for selecting a predetermined value based on correlation scores. Specifically, the system calculates a first correlation score representing the relationship between a first set of data elements and a second correlation score representing the relationship between a second set of data elements. The system then subtracts the first correlation score from the second correlation score to form a difference value. If this difference exceeds a predefined threshold, the second correlation score is selected as the predetermined value. This approach ensures that the predetermined value is dynamically adjusted based on the relative strength of the correlations, improving the accuracy and reliability of subsequent data processing operations. The system may be applied in various domains, including machine learning, data analytics, and decision support systems, where correlation-based analysis is critical.
11. The data processing system of claim 10 , wherein the list is an ordered list, wherein the predetermined value is selected by comparing the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences in correlation scores.
This invention relates to data processing systems that analyze correlation scores between data categories to identify significant changes in relationships. The system generates an ordered list of categories based on their correlation scores with a target variable, then determines a predetermined value by comparing the difference in correlation scores between adjacent categories. The predetermined value is selected when the difference exceeds a given percentage change relative to other differences in the list. This helps identify thresholds where correlation shifts meaningfully, improving data analysis by highlighting key transitions in category relationships. The system may also include a user interface for displaying the ordered list and the predetermined value, allowing users to interact with the analysis results. The method involves calculating correlation scores, ordering the categories, and applying statistical thresholds to detect significant changes, which can be used in applications like predictive modeling, anomaly detection, or trend analysis. The invention enhances data interpretation by automating the detection of meaningful correlation shifts, reducing manual effort and improving accuracy in identifying critical data patterns.
12. The data processing system of claim 9 , wherein the computer program product further comprises: computer usable program code for, prior to receiving the data, training the classification system by using known data collected from known observations and instructing the classification system as to which of the known observations falls into a given classification of the classification system.
This invention relates to a data processing system designed to classify observations using a trained classification system. The system addresses the challenge of accurately categorizing data by leveraging pre-existing knowledge to improve classification performance. The classification system is trained using known data collected from known observations, where the system is explicitly instructed on which observations belong to specific classifications. This training process ensures that the classification system can reliably distinguish between different categories when processing new, unseen data. The system is particularly useful in applications where accurate classification is critical, such as in data analysis, machine learning, and automated decision-making systems. By incorporating a training phase with labeled data, the system enhances its ability to generalize and correctly classify new observations, reducing errors and improving overall efficiency. The invention focuses on the integration of training mechanisms within the data processing system to optimize classification accuracy and reliability.
13. The data processing system of claim 9 , wherein the computer selects the predetermined value either by (1) subtracting a first correlation score from a second correlation score to form a difference, and selecting the second correlation score as the predetermined value when the difference exceeds a given value, or (2) selecting the predetermined value by comparing the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences.
A data processing system analyzes correlation scores between data categories to determine a predetermined value for ranking or filtering purposes. The system compares correlation scores of adjacent categories in a ranked list to identify significant changes. In one approach, the system subtracts a first correlation score from a second correlation score to form a difference and selects the second correlation score as the predetermined value if the difference exceeds a predefined threshold. Alternatively, the system evaluates the difference against other differences in correlation scores between higher-ranked adjacent categories. The predetermined value is chosen when the difference surpasses a specified percentage change relative to these other differences. This method ensures that the selected value represents a meaningful shift in correlation strength, improving the accuracy of data categorization or ranking. The system may apply this technique in various applications, such as recommendation systems, data clustering, or decision-making algorithms, where distinguishing between closely related categories is critical. The approach dynamically adjusts the selection criteria based on the distribution of correlation scores, enhancing adaptability to different datasets.
14. A system on an aircraft configured to classify an input containing data, the system comprising: a processor; a bus connected to the processor; a non-transitory computer readable storage medium connected to the bus; sensors configured to gather the input on or within the aircraft; a tangible input device stored in the non-transitory computer readable storage medium and configured to receive a list of categories from a classification system stored at a computer, wherein each category in the list is assigned a corresponding distinct correlation score that indicates a relevance of a particular category to a particular classification, wherein the distinct correlation score for each category is based on an accumulation of scores for a number of category values for the category; a selection device stored in the non-transitory computer readable storage medium and configured to select a sub-list of categories, wherein the sub-list comprises those categories in the list that have corresponding distinct correlation scores above a predetermined value, and wherein the sub-list of categories comprises less than all of the categories in the list; a data input device stored in the non-transitory computer readable storage medium and configured to receive the input, the input containing the data, wherein the data comprises a snapshot, wherein the data is organized by a plurality of input attributes, wherein each input attribute has a corresponding input category and a corresponding input value, such that the plurality of input attributes has a plurality of input categories and a corresponding plurality of input values, wherein at least some of the plurality of input attributes has corresponding input categories that match at least some of a plurality of categories of the list; a snapshot generator stored in the non-transitory computer readable storage medium and configured to generate a truncated snapshot, the truncated snapshot comprising only attributes from the plurality of input attributes that have corresponding input categories that match categories in the sub-list of categories; and a classifier stored in the non-transitory computer readable storage medium and configured to classify the data using the truncated snapshot and the classification system.
The system is designed for aircraft data classification, addressing the challenge of efficiently processing and categorizing large volumes of sensor data collected onboard. The system includes a processor, a bus, and a non-transitory storage medium connected to the bus, along with sensors that gather input data from or within the aircraft. A tangible input device receives a list of categories from an external classification system, where each category has a distinct correlation score indicating its relevance to a specific classification. These scores are derived from accumulated values for multiple category attributes. A selection device filters this list to create a sub-list of categories, retaining only those with correlation scores above a predetermined threshold, ensuring the sub-list contains fewer categories than the original list. A data input device receives the input data, which is organized into attributes, each with a corresponding category and value. Some of these input categories match categories in the received list. A snapshot generator processes this data to create a truncated snapshot, retaining only attributes that match categories in the filtered sub-list. Finally, a classifier uses this truncated snapshot and the classification system to classify the data. This approach optimizes classification by focusing on the most relevant categories, improving efficiency and accuracy in aircraft data analysis.
15. The system of claim 14 , wherein the selection device is further configured to select the predetermined value by subtracting a first correlation score from a second correlation score to form a difference, and to select the second correlation score as the predetermined value when the difference exceeds a given value.
This invention relates to a system for selecting a predetermined value based on correlation scores. The system addresses the challenge of determining an optimal value from multiple correlation scores by comparing their relative differences. The system includes a selection device that calculates a difference between a first correlation score and a second correlation score. If this difference exceeds a predefined threshold, the selection device selects the second correlation score as the predetermined value. Otherwise, the system may use alternative criteria or default values. The selection process ensures that the chosen value is statistically significant and meets performance requirements. The system may be part of a larger apparatus that processes data, such as in machine learning, signal processing, or data analysis applications. The invention improves decision-making by dynamically adjusting selections based on correlation score comparisons, enhancing accuracy and reliability in automated systems.
16. The system of claim 15 , wherein the list is an ordered list, wherein the selection device is further configured to compare the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences in correlation scores.
This invention relates to a system for categorizing data based on correlation scores, addressing the challenge of accurately determining category boundaries in datasets where distinctions between adjacent categories are subtle. The system generates an ordered list of categories, each associated with a correlation score representing the strength of relationship between the category and the data. A selection device within the system compares the difference in correlation scores between adjacent categories to other differences in the list. The system then evaluates whether this difference exceeds a predetermined threshold, which is dynamically adjusted based on a percentage change relative to other differences in the list. This ensures that category boundaries are set only when the correlation score difference is statistically significant compared to other transitions in the dataset. The system may also include a data processor that generates the correlation scores and a display device that presents the categorized data. The ordered list may be generated by ranking categories based on their correlation scores, and the selection device may apply additional filtering criteria to refine the categorization. The invention improves the accuracy of data segmentation by dynamically adapting the threshold for category separation based on the relative differences in correlation scores.
17. The system of claim 14 further comprising: a trainer stored in the non-transitory computer readable storage medium and configured to, prior to receiving the data, train the classification system by using known data collected from known observations and to instruct the classification system as to which of the known observations falls into a given classification of the classification system.
This invention relates to a machine learning-based classification system for analyzing observational data. The system addresses the challenge of accurately categorizing observations by leveraging pre-trained models and known datasets to improve classification performance. The core system includes a classification module that processes input data to determine its classification based on learned patterns. The system further incorporates a trainer module that pre-trains the classification system using labeled data from known observations. The trainer module ensures the classification system is instructed on which observations correspond to specific classifications, enhancing its accuracy before real-world data is processed. The system is designed to operate on a computing device with a non-transitory storage medium, where the classification and trainer modules are stored and executed. This approach improves the reliability of automated classification tasks by reducing errors in categorization through pre-training with verified datasets. The invention is particularly useful in applications requiring high-precision classification, such as medical diagnostics, quality control, or data analysis.
18. The system of claim 14 , wherein the system further comprises: an associative memory, the associative memory comprising a plurality of data and a plurality of associations among the plurality of data, wherein the plurality of data is collected into associated groups, wherein the associative memory is stored in the non-transitory computer readable storage medium and configured to be queried based on at least indirect relationships among the plurality of data, wherein the data is received in the associative memory, and wherein the selection device is configured to select the sub-list by: the associative memory cycling through all of the plurality of categories, determining whether a given category has a corresponding correlation score above the predetermined value, and including the given category in the sub-list if the corresponding correlation score is above the predetermined value.
The system relates to data processing and retrieval, specifically improving the efficiency of querying and selecting data based on associative relationships. The problem addressed is the difficulty of efficiently retrieving relevant data from large datasets when relationships between data points are complex or indirect. Traditional search methods often fail to capture nuanced connections, leading to incomplete or irrelevant results. The system includes an associative memory that stores a plurality of data items and their interconnections, organized into associated groups. This memory is designed to be queried based on indirect relationships among the data, allowing for more flexible and context-aware retrieval. The associative memory is stored in a non-transitory computer-readable storage medium and processes incoming data to establish and refine these associations. A selection device within the system filters data by cycling through predefined categories, evaluating each category's correlation score against a predetermined threshold. If a category's score exceeds this threshold, it is included in a refined sub-list of results. This approach ensures that only highly relevant categories, determined by their associative strength, are selected, improving the precision of data retrieval. The system enhances search accuracy by leveraging indirect relationships, making it particularly useful in applications requiring deep semantic or contextual understanding, such as knowledge graphs, recommendation systems, or advanced database queries.
19. The system of claim 14 , wherein the list is ordered from highest corresponding distinct correlation score to lowest corresponding distinct correlation score.
A system for organizing and displaying data correlations is disclosed. The system addresses the challenge of efficiently presenting correlation data between multiple data sets, where the relationships may vary in strength and relevance. The system generates a list of correlations between data elements, where each correlation is assigned a distinct correlation score representing the strength or significance of the relationship. The list is then ordered from highest to lowest correlation score, allowing users to quickly identify the most significant correlations. This ordering helps prioritize analysis and decision-making by highlighting the strongest relationships first. The system may also include additional features such as filtering, visualization, or interactive exploration of the correlations. The underlying data sets can be of any type, including numerical, categorical, or time-series data, and the correlations may be computed using statistical, machine learning, or other analytical methods. The ordered list ensures that users can efficiently navigate and interpret the most relevant correlations without being overwhelmed by less significant relationships.
20. The system of claim 14 , wherein the computer selects the predetermined value either by (1) subtracting a first correlation score from a second correlation score to form a difference, and selecting the second correlation score as the predetermined value when the difference exceeds a given value, or (2) selecting the predetermined value by comparing the difference to other differences in correlation scores between adjacent categories higher on the list, and wherein the predetermined value is selected when the difference exceeds a given percentage change in correlation scores relative to the other differences.
This invention relates to a system for analyzing and selecting correlation scores between categories in a ranked list. The system addresses the problem of determining a meaningful threshold or predetermined value to distinguish between adjacent categories in a ranked list, ensuring accurate differentiation based on correlation scores. The system processes correlation scores between categories, which are derived from a method that involves calculating correlation scores for each category pair in the list. The system then selects a predetermined value to serve as a threshold for distinguishing between categories. The selection process involves two approaches. First, the system subtracts a first correlation score from a second correlation score to form a difference and selects the second correlation score as the predetermined value when this difference exceeds a given value. Alternatively, the system compares the difference to other differences in correlation scores between adjacent higher-ranked categories and selects the predetermined value when the difference exceeds a given percentage change relative to these other differences. This ensures that the selected threshold is statistically significant and meaningful in the context of the ranked list. The system enhances decision-making by providing a robust method for distinguishing between categories based on correlation score analysis.
Unknown
March 17, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.